Cell state transition dynamics at single-cell level (Hood. Ozinskv. Huang) Introduction: In metazoa, the very same genome produces a vast diversity of discrete, robust or metastable cell states, as most prosaically manifest in the various cell types of the human body. The switching between cell states is at the core of multicellular development and is also important in diseases, such as cancer, and in regeneration. In this project, we will determine how gene regulatory networks (GRN) regulate the cell state changes, and hence phenotype changes, that occur in development and cancer. Thus, ultimately we want know how these networks specify cell phenotype. This will be achieved by measuring gene expression dynamics at single cell-resolution during cell phenotype change in two clinically relevant settings: (a) cardiomyocytes reprogramming and {b) breast cancer (stem) cell epithelial-mesenchymal transition. Why are single-cell level studies critical for a systems understanding of cell state transitions? One of the key missions in biology is to decode the functional complexity of the human body beyond the traditional molecular profiling of the presumably > 250 distinct cell types. This is highlighted by the finding from single cell analysis that protein or mRNA expression varies by 10-1000-fold between cells even in apparently uniform, clonal cell populations, and that the outliers cells are functionally distinct. Yet few of the measurement techniques that have informed our network models in the past (Project 1) actually achieve single-cell resolution. Standard techniques typically capture only the averaged state of the cellular population, thus blurring unique features of inherent cellular heterogeneity. Conversely, the much needed analysis at the single-cell resolution poses new challenges. Challenges: Conceptual challenges; The emerging capacity to measure gene expression in individual cells and to compute their statistical distribution in an entire cell population confronts us with new fundamental questions: What is the relationship between a pattern of gene expression (a GRN state) and the phenotype for a given cell? Can quantized states (occupied by many cells) be identified for cells exhibiting distinct phenotypes? How does the distribution of gene expression states among cells within a population map into higher levels of organization (developmental trajectory, tissue phenotype)? This project addresses these questions by developing a novel high-throughput measurement work flow that will lead to a new concept of multicellularity that takes population heterogeneity into account. Recent work by members of the Center has revealed that each individual cell in a uniform population is in a distinct state of the very same GRN. Similarly, state transitions need to take into account the multi-step, asynchronous and multi-trajectory nature of cell differentiation, requiring the GRN state to be defined for each cell. Such population heterogeneity must be considered when modeling the dynamics of an entire network as a system (Project 1) or the assembly of cells into multi-cellular structures (Project 2). (See below for the new conceptual framework that addresses this challenge).- Computational challenges: Given the relatively low throughput (<100 variables) the analyses of data produced by the experiments of this project do not pose significant computational challenges. Integrated analysis of network dynamics in single cells is not fundamentally different (after control for noise) from that for cell populations averages, and thus will benefit from the improvements of existing approaches made in Project 1. Since each cell is equivalent to an instance of a network, single cell data provide a greater resolution and affords new insights into the concepts of network robustness studied in Project 1. Technical challenge: In single-cell measurement the technical challenges are three fold: (/) Premeasurement physical handling of individual cells; (//) The distinction between technical variability from inherent biological cell-to-cell variability [given that measurements on the very same cell are not 'aliquots' and cannot be repeated] and (///) Image-based quantitation of phenotype variability. Proposed solutions: (/) Development of new microfluidics-based approaches (complementing flow cytometry) is a main effort in this project. (//) Variability between single cell measurements is a novel problem, related to the new conceptual challenge. We will address it by combining small-scale single-cell measurement with population-level single-cell resolution measurements (flow cytometry) to assess variability due to single-cell sampling. {Hi) Finally, phenotype-based single-cell tools are in place, though the challenge is to couple this to biochemical measurements of the very same cell, as is proposed in the ICM Core research project.